ERRATUM

Development 135, 3473 (2008) doi:10.1242/dev.022053

A dual requirement for Iroquois genes during Xenopus kidney development Pilar Alarcón, Elisa Rodríguez-Seguel, Ana Fernández-González, Ruth Rubio and José Luis Gómez-Skarmeta

There was an error published in Development 135, 3197-3207 in the ePress and the online versions.

On p. 3204, the left- and right-hand columns of text were transposed.

The online issue has now been corrected. The print copy is correct.

We apologise to authors and readers for this mistake. RESEARCH ARTICLE 3197

Development 135, 3197-3207 (2008) doi:10.1242/dev.023697

A dual requirement for Iroquois genes during Xenopus kidney development Pilar Alarcón*, Elisa Rodríguez-Seguel*, Ana Fernández-González, Ruth Rubio and José Luis Gómez-Skarmeta†

The Iroquois (Irx) genes encode evolutionary conserved homeoproteins. We report that Xenopus genes Irx1 and Irx3 are expressed and required during different stages of Xenopus pronephros development. They are initially expressed during mid-neurulation in domains extending over most of the prospective pronephric territory. Expression onset takes place after kidney anlage specification, but before pronephric organogenesis occurs. Later, during nephron segmentation, expression becomes restricted to the intermediate tubule region of the proximal-distal axis. Loss- and gain-of-function analyses, performed with specific morpholinos and inducible wild-type and dominant-negative constructs, reveal a dual requirement for Irx1 and Irx3 during pronephros development. During neurula stages, these genes maintain the specification of the pronephric territory and define its size. This seems to occur, at least in part, through positive regulation of Bmp signalling. Subsequently, Irx genes are required for proper formation of the intermediate tubule. Finally, we find that signalling activates both Irx1 and Irx3 genes in the pronephros.

KEY WORDS: Iroquois, Kidney, Xenopus, Regulation, Patterning

INTRODUCTION the tailbud stage in the intermediate tubule segment of the Studies performed in different vertebrates indicate that most of the pronephros, immediately prior to regionalization of the proximal- genes necessary for pronephros formation in Xenopus are also distal axis (Reggiani et al., 2007). This study further showed that crucial for the formation of more complex mammalian metanephros Irx3, but not Irx1 or Irx2, is required for development of this region or adult kidneys (Carroll et al., 1999; Dressler, 2006; Ryffel, 2003). (Reggiani et al., 2007). Moreover, these similarities at the molecular level correlate with We report that Xenopus Irx1, Irx2 and Irx3 are also expressed physiological homologies. Thus, the tubules of all nephrons have in the pronephric territory during earlier mid-neurula stages. similar subdivisions along the anteroposterior axis with analogous Morpholino loss-of-function analyses, together with distribution of transporters of small molecules and ions along this misexpression of inducible forms of wild-type and dominant- axis (Reggiani et al., 2007; Wingert et al., 2007; Zhou and Vize, negative Irx proteins, reveal a two-step requirement for Irx1 and 2004). This fact, and the accessibility of Xenopus to genetic Irx3 during kidney development. Initially, Irx1 and Irx3 maintain manipulation, makes this animal an excellent model system with the identity of the pronephric territory and define its size. This which to study kidney development. In Xenopus, the specification seems to occur, at least in part, through positive regulation of Bmp of the pronephric anlage occurs in the late gastrula/early neurula signalling. Later, both Irx genes are required for the formation of (stage 12). However, the first sign of pronephric morphogenesis, the the intermediate tubule segment of the pronephros, as reported thickening of the lateral mesoderm, is detected 10 hours later, in the only for Irx3 (Reggiani et al., 2007). In addition, we show that late neurula (stage 20/21). At the tailbud stage (stage 25-30), both genes are regulated by retinoic acid, which is known to be differentiation of the three basic segments occurs: the corpuscle, the necessary to activate early kidney genes and to participate in the tubules and the duct. This is followed by the final maturation of the segmentation of the pronephros in the proximal-distal axis (Cartry organ, which is associated with the physiological specialization of et al., 2006; Wingert et al., 2007). the pronephric tubules along the proximal-distal axis, as observed by the differential activation of several genes encoding different transport proteins (Carroll et al., 1999; Reggiani et al., 2007; Ryffel, MATERIALS AND METHODS 2003; Zhou and Vize, 2004). Plasmid constructions MT-Irx constructs The Iroquois (Irx) genes encode homeoproteins conserved during Constructs were made using Irx alleles from sequences AJ001834, evolution with multiple functions during animal development AJ001835, AF027175, AF338157 and AF338158. MT-Irx1, MT-Irx2 and (Gómez-Skarmeta and Modolell, 2002). Their role during patterning MT-Irx3 have been described previously (de la Calle-Mustienes et al., 2002). of the vertebrate nervous system has been studied in detail (Gómez- To generate MT-Irx4 and MT-Irx5, a fragment from the 5Ј region of each Skarmeta and Modolell, 2002). By contrast, their participation in the cDNA, including unique sites within the ORF (KpnI in Irx4 and ClaI in development of other organs is less well understood. Recently, it has Irx5), were PCR amplified. These sites allowed us to fuse the PCR been reported that Xenopus Irx1, Irx2 and Irx3 are expressed from fragments to the rest of each cDNA. 5Ј primers contained a XhoI (Irx4) or EcoRI (Irx5) site, allowing us to clone these PCRs fragments in frame within the pCS2-MT plasmid (Turner and Weintraub, 1994) at XhoI (MT-Irx4) or Centro Andaluz de Biología del Desarrollo, CSIC/UPO, Carretera de Utrera Km1, between EcoRI and XhoI (MT-Irx5). Primers used were: Irx4, 5Ј- 41013 Sevilla, Spain. ccctcgagATGTCATATCCTCAGTTTGGC-3Ј and 5Ј-GCTCCCATC CA - *These authors contributed equally to this work TGGTACCATACC-3Ј; Irx5, 5Ј-gggaattcaCATGTCCTATCCGCAGGGC- †Author for correspondence (e-mail: [email protected]) 3Ј and 5Ј-ATCCCCTGCATCTCCATC-3Ј. Bold nucleotides indicate restriction sites used for cloning procedures; capitals indicate sequences

Accepted 28 July 2008 present in the cDNAs. DEVELOPMENT 3198 RESEARCH ARTICLE Development 135 (19)

Irx-MT constructs images taken at the same magnification, we used the histogram function of To generate Irx-MT constructs, a fragment from the 3Ј region of each cDNA, Photoshop to compare the size of the area expressing different markers in including unique sites within the ORF (SacI in Irx1, PstI in Irx2, SacII in the injected versus the uninjected sides of the same embryo. Grafts were Irx3, KpnI in Irx4 and XbaI in Irx5 cDNAs), were PCR amplified. For Irx1, performed as previously described (Gómez-Skarmeta et al., 1999). Irx2, Irx3 and Irx5, the fusion to the rest of each cDNA in pCS2-MT was carried out as follows: EcoRI/SacI, PstI, SacII or XbaI fragments that RESULTS contain the 5Ј cDNA regions of Irx1, Irx2, Irx3 or Irx5, respectively, were cloned into pBluescript. These fragments were excised with HindIII, SacI, Expression patterns of Irx genes during Xenopus PstI, SacII or XbaI and ligated with the PCR fragment in pCS2-MT between pronephros development the HindIII and ClaI (Irx1) or BamHI (Irx2, Irx3 and Irx5) sites. The full Irx We examined the expression of the full complement of Xenopus Irx ORF in frame with the tag was then transferred into the EcoRI site of genes during pronephros development. Irx1 and Irx2 were largely pCS2+. For Irx4, a ClaI/KpnI fragment containing the 5Ј cDNA was ligated co-expressed during pronephric development from late neurula with the corresponding 3Ј PCR fragment, expanding the 3Ј cDNA into the onwards (Fig. 1A-H). Irx3 was temporally and spatially expressed ClaI site of pCS2-MT. Primers used were: for Irx1, 5Ј-GCAACAA - in similar, but not identical, territories (Fig. 1I-L). Irx4 was only GCCCAGATGG-3Ј and 5Ј-ccaatcgatGGCAGAGG GA AGTGCTG-3Ј; for found in the pronephros at tadpole stage, Irx5 was never detected in Irx2, 5Ј-GCCGACCATCTTTGCG-3Ј and 5Ј-ggggatccTGGGTATGG - the kidney territory (see Fig. S1 in the supplementary material) and Ј Ј Ј Ј TTGTACTCC-3 ; for Irx3, 5 -CACAGCC CCATGTTCTGG-3 and 5 - Irx6 was not expressed at these stages (de la Calle-Mustienes et al., ggggatccGGATGAGGATAAAGCGGA-3Ј; for Irx4, 5Ј-CCATGGTACCT - ACCCTCG-3Ј and 5Ј-ccaatcgatAGCA AGATGTTCTGTTCCT-3Ј; for 2005). Irx1, Irx2 and Irx3 were initially expressed at mid neurula Irx5, 5Ј-CTTCTCCATCTAGATCTCC-3Ј and 5Ј-ggggatccAATGCTAG - stages in the pronephric anlage (Fig. 1A,E,I). Double staining with AC ATACCTTTCTTC-3Ј. Bold nucleotides indicate restriction sites used the early pronephric markers Lim1 and Pax8 (Carroll and Vize, for cloning procedures; capitals indicate sequences present in the cDNAs. 1999; Heller and Brandli, 1999) confirmed that Irx1 and Irx3 were expressed in the pronephric field (insets in Fig. 1A,I; Fig. 1M,Q). MT-Irx-GR constructs Moreover, although the Irx3 pronephric domain was broad and We first generated Irx-MT-GR derivatives by cloning the GR domain within the XhoI and XbaI fragment located 3Ј of the Irx-MT in the pCS2 Irx-MT encompasses most, if not all, of the Lim1 territory, Irx1 was confined vectors. The hormone-inducible GR domain was obtained by PCR using the to the dorsal area of the Lim1 field. At late neurula stage, Irx1 was oligonucleotides 5Ј-cccctcgagATCCCCTCTGAAAATCC-3Ј and 5Ј- still confined to the dorsal pronephric domain (Fig. 1B,N). By ctctagaCACTTTTGATGAAACAGAAG -3Ј from a MyoD-GR plasmid contrast, Irx3 became now restricted to the ventral pronephric kindly donated by H. Sive. To make the chimeric mRNAs of these constructs territory (Fig. 1J,R). During tailbud stages, Irx1 expression became insensitive to the MOs, we introduced a MT 5Ј by fusing these constructs localized in the intermediate tubule, as judged by double staining with their corresponding MT-Irx as follows: for Irx1, a 5Ј EcoRI-SacI with Pax8 (Fig. 1C, inset) or with the proximal tubule marker Sglt1k fragment from MT-Irx1 was ligated to a 3Ј SacI-NotI fragment from MT- (Fig. 1O) (Reggiani et al., 2007; Zhou and Vize, 2004). At this stage, Irx1-GR in pCS2-MT; for Irx2, a 5Ј EcoRI-ApaI fragment from MT-Irx2 was Irx3 also became confined to the intermediate tubule, but its ligated to a 3Ј ApaI-NotI fragment from MT-Irx2-GR in pCS2-MT; for Irx3, expression extended into the distal area of the proximal tubule and a 5Ј EcoRI-SacII fragment from MT-Irx3 was ligated to a 3Ј SacII-NotI fragment from MT-Irx3-GR in pCS2-MT. into the distal tubule, as determined by double staining with Pax8 (Fig. 1K, inset) or with the intermediate tubule marker Nkcc2 (Fig. In situ hybridization, X-Gal and antibody staining 1S) (Reggiani et al., 2007; Zhou and Vize, 2004). The expression Antisense RNA probes were prepared from cDNAs using digoxigenin or patterns of Irx1 and Irx3 were maintained at tadpole stages (Fig. fluorescein labels (Roche). Xenopus specimens were prepared, hybridized 1D,L,P,T). In addition, late Irx1 expression was also observed in and stained as described (Harland, 1991). X-Gal staining was performed migrating ventral mesoderm (Fig. 1D, blue arrowhead). We accordingly to Coffman et al. (Coffman et al., 1993). Double fluorescent in conclude that Irx genes have dynamic patterns of expression during situ hybridization was performed as previously described (Zhou and Vize, 2004). Antibody staining was performed as previously described (Gómez- pronephros development and that their expression in the pronephric Skarmeta et al., 2001). Monoclonal antibodies 3G8 and 4A6 were kindly territory starts much earlier than recently reported (Reggiani et al., provided by E. Jones. The monoclonal antibody 12/101, generated by J. P. 2007). Brockes, was obtained from the Developmental Studies Hybridoma Bank (NICHD and The University of Iowa, Department of Biological Science, Loss of Irx1 and Irx3 function impairs pronephros Iowa City, IA 52242). development To examine the requirement for Irx genes during Xenopus pronephros In vitro RNA synthesis, microinjection of mRNA and morpholinos, development, we interfered with the activity of each Irx mRNA by and grafts DNAs were linearized and transcribed as described (Harland and Weintraub, injecting specific translation-blocking morpholinos (MOs). As 1985) with GTP cap analogue (New England Biolabs). SP6, T3 or T7 RNA Xenopus laevis is pseudotretaploid, we identified all ESTs available polymerases were used. After DNAse treatment, RNA was extracted with in the database for each Irx gene and found one allele for Irx4 and two phenol-chloroform, column purified and precipitated with ethanol. mRNAs alleles for Irx1, Irx2, Irx3 and Irx5. We designed specific Irx MOs that for injection were resuspended in water. Synthetic mRNAs or MOs were block translation from both Irx alleles when present in the genome. injected into V2.2 blastomeres with 2-4 nl solutions. The following The specificity of these MOs is shown in Fig. S2 in the supplementary morpholinos were used in this study: MOIrx1, 5Ј-CATGTC TCTC - material. Embryos injected with any of these Irx MOs showed CGGCAGGGAAATCGC-3Ј; MOIrx2, 5Ј-AGGTAACCCTGAG GATA - different degrees of neural defects (E.R.-S., P.A. and J.L.G.-S., GGACATGG-3Ј; MOIrx3, 5Ј-CTGTGGGAAGGACATGGTGCAGCCG- unpublished) indicating that they are effective in blocking the activity 3Ј; MOIrx3.2, 5Ј-AGCTGTGGGAAGGACATGGTGCAGC-3Ј; MOIrx4, of their respective Irx genes. To reduce the MO effects on off-target 5Ј-GTAGCCAAACTGAGGATATGACATT-3Ј; and MOIrx5, 5Ј- CAAGTAGCCCTGCGGATAGGACATG-3Ј. MOIrx1 and MOIrx5 are tissues, in all experiments, we targeted the pronephric anlage by 100% homologous to the Irx1 and Irx5 alleles used in this study. The second injecting the V2.2 blastomere of 8-16-cell stage embryos. We then Irx5 and Irx1 alleles contain 1 and 2 sequence mismatches, respectively, with evaluated the effect these injections on the early renal markers Lim1 their corresponding MOs. The other Irx MOs have 100% homology with all and Pax8. We considered that an embryo had an altered pronephros

their corresponding Irx alleles. In the MO- or mRNA-injected embryos, in when the area expressing the corresponding marker on the injected DEVELOPMENT Irx genes in Xenopus kidney development RESEARCH ARTICLE 3199

Fig. 1. Expression patterns of Xenopus laevis Irx genes during pronephros development. Embryos are shown in lateral views (except when indicated); red arrowheads indicate the kidney territory. (A-D) Expression pattern of Irx1 at indicated developmental stages. At mid- (A) or late (B) neurula, Irx1 is detected in the dorsal pronephric territory. Insets show pronephric territory of an embryo double-stained for Lim1 or Pax8 (blue) and Irx1 (purple). Irx1 expression is restricted dorsally. During tailbud (C) or tadpole (D) stages, Irx1 expression shifts to a more ventral region that will form the intermediate tubule. Inset in C indicates an embryo double-stained for Pax8 (blue) and Irx1 (purple). Note the ventral position of Irx1 in the future intermediate tubule. Inset in D indicates a higher magnification of the pronephric Irx1 territory. Note the expression of Irx1 in the migrating ventral mesoderm (blue arrowheads). (E-H) Irx2 shows an expression pattern similar to that of Irx1, although it is not expressed in ventral migrating mesoderm. (I-L) Spatial distribution of Irx3 mRNA. (I) At mid-neurula, Irx3 mRNA is detected in a broad domain that contains most of the pronephric territory. Inset indicates pronephric territory of an embryo double-stained for Lim1 (blue) and Irx3 (purple). (J) At late neurula/early tailbud stages, Irx3 becomes restricted to the ventral pronephric territory. Inset indicates Irx3 ventral restriction in an embryo co-stained for Pax8 (blue). (K,L) From tailbud stages, Irx3 expression is detected in the intermediate tubule. Inset in K indicates embryo double-stained for Pax8 (blue) and Irx3 (purple). Inset in L indicates high magnification of the pronephric Irx3 territory. (M,N) Lateral view (M) and transverse section (N) of late neurula embryos showing Lim1 (blue) and Irx1 (purple) expression. Irx1 expression is restricted to the dorsal pronephric anlage. (O,P) Double staining for Sglt1k (green) and Irx1 (red) in tailbud (O) or tadpole (P) embryos. Insets indicate single Irx1 red channel. The Irx1 expression domain is located just distal to that of Sglt1k. (Q,R) Lateral view (Q) and transverse section (R) of late neurula embryos showing Lim1 (blue) and Irx3 (purple) expression. There is initial broad expression of Irx3 in most of the pronephric anlage (Q) and a later restriction to the ventral pronephros (R). (S,T) Double staining for Nkcc2 (green) and Irx3 (red) in tailbud (S) or tadpole (T) embryos. Insets show single Irx3 red channel. The expression domains of both genes largely overlap, but the Irx3 domain extends proximally into the proximal tubule, whereas Nkcc2 extends distally into the distal tubule. side varied by more than 20% relative to the uninjected control side. the muscle-specific antibody 12/101, we found that this was not the Differences greater that 20% between the two pronephros of a single case (Fig. 2A,B,I,J). We also monitored the effect of injecting MOIrx1 embryo were very rarely observed in non-injected embryos or in or MOIrx3 on the expression of genes expressed at tadpole stages, embryos injected with a control MO (<2%, n=83). Injection of 8 ng during the maturation of the pronephros. All three genes examined, of MOs against Irx1 or Irx3 (Fig. 2), but not Irx2, Irx4, Irx5 or a Sglt1k, Nkcc2 and Nbcc1 [which are expressed in proximal, control MO (Fig. S3 in the supplementary material; not shown), intermediate and distal tubule, respectively (Reggiani et al., 2007; caused renal defects. Thus, at mid-late neurula stage, the territory Zhou and Vize, 2004)], were downregulated in MOIrx1 (38-47%, expressing Lim1 or Pax8 was reduced in most Irx1 or Irx3 morphant n=34-39) or MOIrx3 (51-96%, n=24-41) morphants (Fig. 2C-H,K-P). embryos (Fig. 2A,B,I,J, and not shown). The average of pronephros The reduction of Irx function did not significant altered the rate of cell size reduction was around 40-50%, and was observed in 57% (n=159) proliferation or cell death in the kidney territory (not shown). and 83% or (n=74) of the MOIrx1- and MOIrx3-injected embryos, We next examined the effect of knocking-down simultaneously respectively. As muscles are a source of signals that influence kidney both Irx1 and Irx3 (with a mix containing 3-4ng of each MO). Co- development (Seufert et al., 1999), we determined whether, in the Irx injection of Irx1 and Irx3 MOs at half doses caused phenotypes similar

morphant embryos, muscle development was altered. By staining with to individual MO injections at double concentration. The DEVELOPMENT 3200 RESEARCH ARTICLE Development 135 (19)

Fig. 2. Irx1 and Irx3 are necessary for kidney formation in Xenopus. Embryos are shown in lateral views; red arrowheads indicate the kidney territory. Embryos were injected in a single blastomere (V2.2) at the 8- to 16-cell stage and lacZ mRNA was used as linear tracer. Control and injected sides of the same embryo are shown, respectively, of the same specimen. The gene examined in each condition is indicated in the right upper corner of the panels in all figures. (A-H) Embryos injected with MOIrx1 showed reduced Lim1 expression at late neurula (A,B) and downregulation of Sglt1k (C,D), Nkcc2 (E,F) and Nbcc1 (G,H) expression at tadpole stages. Inset in (A) indicates a transverse section of the embryo shown in the major panel. (I-P) Similar results were found upon MOIrx3 injection.

simultaneous impairment of Irx1 and Irx3 resulted in the loss of Lim1 and 12/101 (Fig. 3G-J). Interestingly, in these Irx-depleted embryos, (58%, n=66) and Pax8 (89%, n=57) expression only at late neurula the number of ventral muscle fibres was reduced (Fig. 3G,H, blue stage, coinciding with the onset of expression of Irx1 and Irx3, but it arrowheads). This suggests that the expression of Irx1 in ventral did not affect the expression of these genes at early neurula (Fig. 3A- migrating muscle cells may be required for proper development of F). In these injected embryos, the differentiated phronephric structures these muscles. Sections through these injected embryos suggested that at tadpole stages, but not the somitic muscles, were severely reduced, cells that lose their kidney fate are likely to end up as fibroblasts, as an as determined by triple staining with the antibodies 3G8 and 4A6, increased number of cells with fibroblast shape are detected in the Irx which label the tubules and the duct, respectively (Vize et al., 1995), depleted side (Fig. 3J).

Fig. 3. Irx genes are not required for the initial activation of the early kidney genes. Embryos are shown in lateral views (except A and B, which are dorsal views); red arrowheads indicate the kidney territory. Embryos were injected in a single blastomere (V2.2) at the 8- to 16-cell stage. Xenopus embryos injected with a mix of Irx1 and Irx3 MOs and lacZ mRNA were assayed for the expression of Pax8 and Lim1 genes at early (A,B) or late (C-F) neurula stages. (A,B) Impairment of Irx gene function does not affect early expression of Pax8 (A) or Lim1 (B). (C-F) By contrast, depletion of Irx activity downregulates the expression of these genes at late neurula stage. (G,H) Tadpole embryos injected with Irx1 and Irx3 MOs and triple labelled for muscle (12/101, brown), pronephric tubules (3G8, blue) and duct (4A6, purple). The injected side (H) shows strong impairment of kidney tissue (red arrowheads) and reduced number of ventral muscle fibres (blue arrowheads) when compared with the control side (G). (I) Transverse section of the embryo shown in H. (J) The same section after treatment with propidium iodide. An increased number of fibroblast-like cells in the injected right side compared with the control left side (arrowheads). DEVELOPMENT Irx genes in Xenopus kidney development RESEARCH ARTICLE 3201

Fig. 4. Overexpression of Irx genes in Xenopus expands the pronephric territory. Embryos are shown in lateral views (except when indicated). Embryos were injected in a single blastomere (V2.2) at the 8- to 16-cell stage and lacZ mRNA was used as linear tracer. Neurula (A-P) or tadpole (Q-R) embryos injected with different mRNAs. Transverse sections of tadpole embryos are shown in S,T. (A-H) Overexpression of 300 pg of MT-Irx1-GR (A-D) or MT-Irx3-GR (E-H) mRNAs expands (arrowheads) ventrally the expression of Lim1 (A,B,E,F) and Pax8 (C,D,G,H) upon addition of dexamethasone (Dex) at stage 14, whereas no effect was observed in the absence of Dex (not shown). (I-P) Embryos co-injected with a mix of Irx1 and Irx3 MOs and MT-Irx3-GR mRNAs show strong downregulation of Lim1 (I,J; arrowhead) and Pax8 (K,L; arrowhead) in the absence of Dex. (M-P) This phenotype is rescued upon addition of hormone at stage 14. (Q-T) Tadpole embryos injected with MT-Irx1-GR mRNAs and Dex treated at stage 14 show enlarged kidneys (Q,R; arrowhead), as determined by staining with the 3G8 antibody. (S) Transverse section of a similarly injected embryo. (T) The same section treated with propidium iodide for nuclear staining. The control and the enlarged pronephros show the same cellular morphology.

The autonomous requirement of Irx1 and Irx3 for pronephros the MOs (see Materials and methods). These constructs allowed formation was further supported by a transplantation experiment. A the induction of Irx function after gastrulation, thus eliminating graft of lateral plate from a late gastrula embryo co-injected with possible earlier effects of Irx genes on mesoderm formation Irx1MO, Irx3MO and GFP mRNA was transplanted to the equivalent (Glavic et al., 2001). All three MT-Irx-GR proteins behaved area of a wild-type host. In the transplanted embryo, the expression of similarly in overexpression studies (see below and not shown). Lim1, but not that of neural or muscle markers ( and 12/101, Consistent with a requirement for Irx genes during pronephric respectively), was impaired (see Fig. S4 in the supplementary development, overexpression of MT-Irx1-GR or MT-Irx3-GR material) (66%, n=6). This was not observed after transplantation of mRNAs, upon addition of dexamethasone (Dex) at stage 14, a control graft from an embryo injected with only GFP mRNA (100%, triggered a ventral expansion of Lim1 and Pax8 (Fig. 4A-H; 60% n=7; not shown). Finally, we also monitored the expression of several of the embryos showing enlarged pronephros, n=132). In most additional kidney markers (Osr2, Nhf1β, Wnt4 and Wt1) in double embryos, the pronephros at the injected side was about 50% larger Irx1 and Irx3 morphant embryos. Expression of all genes was reduced than the pronephros at the control non-injected side. This (see Fig. S5 in the supplementary material). Together, our results expansion enlarged the differentiated kidney tissue, as determined indicate that Irx1 and Irx3 are activated following the specification of by staining with the tubules antibody (3G8) (Fig. 4Q-T). We next the kidney anlage and are autonomously required for the maintenance determined the ability of these MT-Irx-GR constructs to rescue of this specification. the defects observed in Irx morphant embryos. Although interference with Irx1 and Irx3 function with a mix of MOs Gain of Irx1 and Irx3 function expand the caused downregulation of Lim1 and Pax8 (Fig. 4I-L; Fig. 3C-F), pronephric field co-injection of Irx MOs with MT-Irx1-GR or MT-Irx3-GR Our results prompted us to test the effect of the misexpression of mRNAs rescued the expression of these genes, but only upon Irx genes. We first generated hormone-inducible forms of the hormone addition at early neurula stage (Fig. 4M-P and not

Irx1, Irx2 and Irx3 proteins (MT-Irx-GR) that are insensitive to shown; 15% reduced and 55% enlarged pronephros, n=100). DEVELOPMENT 3202 RESEARCH ARTICLE Development 135 (19)

Fig. 5. Early Irx gene requirement for pronephros development occurs at neurula stages. Embryos are shown in lateral views. Embryos were injected in a single blastomere (V2.2) at the 8- to 16-cell stage and lacZ mRNA was used as linear tracer. Late neurula-early tailbud Xenopus embryos co-injected with 500 pg of HD-E1A-GR (A-H), HD-GR (I-P) or HD-EnR-GR (I,M, inset) mRNAs and assayed for expression of Lim1. (A-H) Embryos injected with a hormone-inducible activating form of Irx (HD-E1A-GR) show expanded Lim1 only when Dex was added during mid neurula stages (arrowheads). (I-P) Embryos injected with a hormone-inducible dominant negative (HD-GR) form of Irx show reduced Lim1 expression only when Dex was added during mid neurula stages (arrowheads). The same results were found with a hormone-inducible repressing form of Irx (HD-EnR-GR) (I and M, inset and not shown).

These results indicate that the pronephric expression of Irx genes is required to maintain the transcription of the key renal genes Lim1 and Pax8, and to define the size of the pronephric anlage.

During neurula stage Irx proteins act as activators in the pronephric field To cast light on the way Irx proteins function during these processes, we overexpressed different hormone-inducible constructs with the homeodomain (HD) of Irx1 alone or fused to either the E1A activator or the (EnR) repression domains (HD-GR, HD-GR-E1A and HD-GR-EnR). It is known that, during neural development, Irx proteins act as transcriptional repressors, as overexpression of wild-type Irx proteins or HD-GR- EnR fusions cause similar phenotypes, whereas HD-GR and HD- GR-E1A behave as dominant-negative molecules (Gómez- Skarmeta et al., 2001). By contrast, during kidney development, overexpression of HD-GR-E1A mRNA (Fig. 5A-H) mimicked the ventral expansion of Lim1 (64% of the embryos with enlarged pronephros, n=48) caused by wild-type Irx mRNAs (Fig. 4). Conversely, the overexpression of HD-GR (Fig. 5I-P) or HD-GR- EnR (inset in Fig. 5I, M, and not shown) mRNAs promoted downregulation of Lim1 (88% of the embryos showing reduced Fig. 6. Irx gene loss of function kidney defects are partially pronephros, n=54). Therefore, during pronephros development, rescued by increased Smad1 activity in Xenopus. Embryos are HD-GR and HD-GR-EnR proteins behave as dominant-negative shown in lateral views and red arrowheads indicate the kidney territory. molecules that interfere with Irx function. Similar results were Embryos were injected in a single blastomere (V2.2) at the 8- to 16-cell found when Pax8 expression was examined (not shown). In stage and lacZ mRNA was used as linear tracer. (A,B) Injection of addition, by providing Dex at different stages of development, we 500 pg of Smad1GR mRNA, upon addition of dexamethasone (Dex) at stage 14, expanded Pax8 expression. No effect was observed in the also found that the requirement for Irx function during pronephros absence of hormone (not shown). (C-F) In embryos co-injected with development occurred around mid neurula (stages 15-17). 500 pg of Smad1GR mRNA and 4 ng of each Irx1 and Irx3 MOs Pax8 Overexpression of Irx proteins at later stages (20-22) had little expression was downregulated (C,D) or rescued (E,F) in the absence or effect on Lim1 and Pax8 (Fig. 5). These data suggest that Irx presence of Dex, respectively. (G,H) Depletion of Irx1 and Irx3 impaired

proteins act as transcriptional activators to maintain the kidney Bmp7 expression. DEVELOPMENT Irx genes in Xenopus kidney development RESEARCH ARTICLE 3203

Fig. 7. Injection of different doses of Irx1 or Irx3 MOs reveal an early and a late requirement of this gene during pronephric development in Xenopus. Embryos are shown in lateral views and red arrowheads indicate the kidney territory. Embryos were injected in a single blastomere (V2.2) at the 8- to 16-cell stage and lacZ mRNA was used as linear tracer. (A-D) Injection of low doses (4 ng) of Irx1 MO had little effect on Sglt1k expression (A,B) but downregulated the proximal domain of Nkcc2 (C,D). (E-L) Injection of low doses (4 ng) of two different Irx3 MOs downregulated the distal expression of Sglt1k (E,F,I,J) and the proximal domain of Nkcc2 (G,H,K,L). No effect was observed in the duct, as determined by Gata3 expression (E,F,I,J). (M-P) Injection of high doses (8 ng) of MOIrx3.2 strongly downregulated Sglt1k (M,N) and Nkcc2 (O,P). Most injected embryos were malformed, as shown in O,P. Nevertheless, a few displayed normal morphology, like that shown in M,N.

anlage properly and to define the size of this territory, and confirm Irx1 and Irx3 genes are required for proximal- that they are initially required before pronephros morphogenesis distal patterning of the pronephros takes place. Irx3, but not Irx1, has been shown to be required for formation of the intermediate tubule segment of the pronephros (Reggiani et Rescue of Irx-dependent kidney defects by al., 2007). We examined whether Irx1 was also required in this increasing Bmp signalling late process. Injection of high doses of Irx MOs downregulated The Bmp pathway is implicated at several steps during vertebrate all proximal-distal genes, probably because of the early kidney development (Cain et al., 2008). A recent report showed requirement of Irx genes for maintaining the kidney anlage, thus that blocking this pathway during Xenopus neurula stages preventing examination of later Irx functions. To try to overcome impaired pronephros formation (Bracken et al., 2008). These this problem, we partially reduced Irx function by injecting Irx1 results resemble those observed by reducing Irx function. As Irx and Irx3 MOs at lower doses. We complement these experiments genes modulate Bmp signalling during neural development injecting a second Irx3 MO (MOIrx3.2) that, in a previous report, (Gómez-Skarmeta et al., 2001), a similar situation may occur was unable to reveal an early Irx3 requirement (Reggiani et al., during kidney development. To evaluate this, we increased Bmp 2007). Blastomere injection of 4 ng of Irx1 MO or either one of signalling at neurula stages, by overexpressing an inducible the two Irx3 MOs caused little effect on Lim1 and Pax8 Smad1 construct (Smad1GR) in the absence of Irx1 and Irx3 expression in late neurula or tailbud stages (not shown). By activity. Induction of Smad1GR at stage 14 caused an expansion contrast, at tadpole stage, embryos injected with Irx1 or with any of kidney territory (51%, n=35), as determined by Pax8 of the Irx3 MOs showed downregulation of the proximal domain expression (Fig. 6A,B). In the absence of Dex, embryos co- of Nkcc2 (40-50%, n=19-26) (Fig. 7C,D,G,H,K,L). This was also injected with Smad1GR mRNA and Irx1 and Irx3 MOs showed accompanied by a reduction in the distal expression of Sglt1k in downregulation of Pax8 (Fig. 6C,D; 41%, n=51). This phenotype the MOIrx3, but not in the MOIrx1, injected embryos (Fig. is partially rescued by Dex treatment at stage 14 (Fig. 6E,F; 21% 7A,B,E,F,I,J). These results are consistent with the expression with Pax8 downregulated, n=48). Thus, part of Irx function seems domains of Irx1 and Irx3, and suggest that both genes are required to be to positively modulate the Bmp pathway. Bmp7 is expressed for proximal-distal patterning, as was already reported for Irx3 and required for kidney development (Dudley et al., 1995; Luo et (Reggiani et al., 2007). We also tried to address why, in the al., 1995; Wang et al., 1997). We examined whether its expression previous report (Reggiani et al., 2007), an early requirement for depended on Irx activity. Indeed, as for other kidney markers, in Irx3 was not detected. For that, we injected one of the Irx3 MO double Irx1 and Irx3 morphant embryos Bmp7 expression was they used (MOIrx3.2) at higher (8 ng) doses. At this

downregulated (Fig. 6G,H). concentration, most embryos injected with our Irx3 MO were DEVELOPMENT 3204 RESEARCH ARTICLE Development 135 (19)

Fig. 8. Temporal requirement for Irx function during proximal-distal pronephric patterning. Embryos are shown in lateral views and red arrowheads indicate the kidney territory. Embryos were injected in a single blastomere (V2.2) at the 8- to 16-cell stage and lacZ mRNA was used as linear tracer. (A-H) Xenopus embryos injected with HD-GR mRNA. (A-D) Impairment of Irx activity during neurula stages downregulated Sglt1k, Gata3 (A,B) and Nkcc2 (C,D) expression. (E-H) Impairment of Irx activity during tailbud stages did not affect Sglt1k or Gata3 (E,F) but reduced Nkcc2 (G,H) expression. (I-P) Embryos injected with MT- Irx1-GR mRNA. (I-L) Increasing Irx1 function during neurula caused ectopic, patched Sglt1k (I,J) and enlarged Nkcc2 expression domains (K,L). No effect was observed on the duct marker Gata3 (I,J). (M-P) Increasing Irx1 function during tailbud did not affect Sglt1k or Gata3 (M,N) but enlarged Nkcc2 (O,P) expression. (Q-X) Similar results were found in embryos injected with MT-Irx3-GR mRNA.

healthy and showed the strong reduction of all segment markers Retinoic acid regulates pronephric expression of shown in Fig. 2. By contrast, the majority of the embryos injected Irx1 and Irx3 with 8 ng of MOIrx3.2 were malformed. These malformed Retinoic acid (RA) is a requisite for the activation of early kidney embryos also lost most markers (Fig. 7M-P). This might explain genes and for the late segmentation of the pronephros (Cartry et al., the discrepancy if those embryos with stronger phenotypes and 2006; Wingert et al., 2007). Therefore, RA may regulate Irx genes malformations were not taken into account in the previous report during kidney development. To test this, we examined Irx (Reggiani et al., 2007). expression in embryos injected with 100 pg of mRNAs encoding To further examine the Irx requirement in pronephric proximal- either a dominant-negative (RAR-DN) or a constitutively active distal patterning, we injected inducible wild-type or dominant- (RAR-Vp16) RA (Blumberg et al., 1997). Embryos with negative Irx constructs, activated them at different developmental reduced or increased RA signalling showed down or upregulation, stages and examined their effect on pronephric proximal-distal respectively, of the expression of Irx genes in the kidney (Fig. 9A- patterning. A dominant-negative Irx construct induced at stage 13 H). Thus, RA positively regulates Irx genes in the pronephros. We downregulated Sglt1k, Nkcc2 and Gata3 (84-90%, n=19-20) (Fig. then determined when RA signalling is required for kidney Irx 8A-D), markers of proximal tubule, intermediate tubule and duct, expression. To this end, we incubated Xenopus embryos at different respectively (Reggiani et al., 2007; Wingert et al., 2007; Zhou and developmental stages (12.5, 15 or 25) for 1 hour with a control Vize, 2004). By contrast, the same construct induced at stage 27 solution (DMSO), with 4-diethylaminobenzaldehyde (DEAB 30 downregulated Nkcc2 (59%, n=17), whereas Sglt1k and Gata3 µM), the chemical inhibitor of the RA producing enzyme Raldh2 or were not affected (Fig. 8E-H). These results are consistent with with RA (10 µM). Reducing or increasing RA signalling, down or an early and a late requirement for Irx function. A further upregulated, respectively, the expression of both genes, but only confirmation of this dual function was obtained by examining when the drug treatments were done at late gastrula stage (12.5) these proximal-distal markers in embryos injected with Dex- (Fig. 9I-Y). inducible Irx1 or Irx3 proteins. Incubation of injected embryos with Dex from stage 13 caused ectopic patches of Sglt1K (38- DISCUSSION 47%, n=15-24) (Fig. 8I,J,Q,R) and an enlarged Nkcc2 domain The development of the pronephros can be subdivided in three major (52-60%, n=17-24) (Fig. 8K,L,S,T). No clear effect was found on steps: specification of the anlage, morphogenesis of the nephron and Gata3 expression (Fig. 8I,J,Q,R). Addition of hormone at stage the generation of different proximal-distal territories. Here, we show 27, expanded the Nkcc2 expression domain (50-60%, n=14-22) that the homeodomain genes Irx1 and Irx3 play essential functions (Fig. 8O,P,W,X), but did not affect Sglt1K expression (Fig. in two of these steps: maintenance of the specified anlage and

8M,N,U,V). segmentation of the nephron. DEVELOPMENT Irx genes in Xenopus kidney development RESEARCH ARTICLE 3205

Fig. 9. Irx1 and Irx3 are positively regulated by retinoic acid signalling. Embryos are shown in lateral views and red arrowheads indicate the kidney territory. Embryos were injected in a single blastomere (V2.2) at the 8- to 16-cell stage and lacZ mRNA was used as linear tracer. All panels show Xenopus tadpole embryos. (A,B,E,F) Embryos injected with a dominant negative RA receptor mRNA (RAR-DN) showed impaired Irx1 (A,B) and Irx3 (E,F) expression. (C,D,G,H). Embryos injected with a constitutive RA receptor mRNA (RAR-VP16) showed a strong expansion of Irx1 (C,D) and Irx3 (G,H) expression. (I-Y) Embryos treated at different developmental stages (as indicated) with DMSO (I,M,Q,U), with the inhibitor of RA signalling pathway DEAB (J-L,R-T) or with RA (N-P,V-Y), and analyzed for Irx1 (I-P) or Irx3 (Q-Y) expression. Both genes negatively or positively responded to DEAB (J,R) or RA (N,V), respectively, only when the drugs were added at stage 12.5.

Irx genes are required to maintain the kidney Consistent with their initial activation in the pronephric field anlage before pronephros morphogenesis during neurulation, Irx function is dispensable for the initial Irx1 and Irx3 show dynamic expression patterns during pronephros activation of the early kidney determinants Ors1, Ors2, Pax8 and development. We find that Irx genes are initially expressed in the Lim1 that occurs at late gastrula (Carroll and Vize, 1999; Heller and pronephric territory at neurula stage. This occurs after the initial Brandli, 1999; Tena et al., 2007). However, depletion of Irx1 or Irx3 specification of this territory by Ors, Lim1 and Pax8 genes at impairs the expression of all kidney genes examined at tailbud and gastrula stage, but before morphological or molecular signs of tadpole stages. This Irx function seems to be autonomous, as kidney morphogenesis at late neurula-early tailbud sages. Irx1 is downregulation of kidney genes occurs without affecting neural or initially activated in the dorsal pronephric territory. By contrast, Irx3 other mesodermal tissues. Consistent with a requirement for Irx is initially expressed in most of the kidney anlage, but it becomes genes before pronephros morphogenesis, time-controlled loss or confined to the ventral pronephros territory. This dorsal-ventral gain of Irx function during neurula, but not during tailbud stages, subdivision of the prospective kidney may reflect the initial reduces or expands the pronephric field, respectively. Interestingly, subdivision of the pronephric territory by Notch signalling into a gain of Irx function expands but does not promote ectopic dorsal region that will generate glomus and proximal tubule, and a expression of Xlim1 and Pax8, as found when Osr genes or Pax8 and ventral domain that will give rise to distal tubule and duct Xlim1 are overexpressed (Carroll and Vize, 1999; Tena et al., 2007). (McLaughlin et al., 2000; Taelman et al., 2006). However, we have This suggests that Irx genes alone are unable to trigger the kidney not detected an alteration of Irx1 or Irx3 expression by manipulating program. Our results indicate that Irx genes are expressed and Notch signalling (not shown), which suggests that Notch does not required before appearance of any sign of pronephros regulate Irx genes. Interestingly, the expression of Irx genes slightly morphogenesis. This early Irx gene requirement for kidney precedes the onset of expression of Dl1 and Notch1, which is development is likely to be conserved in other vertebrates as these compatible with Irx genes participating in the regulation of Notch genes are also expressed in the early kidney anlage of other signalling in the pronephros. Indeed, Irx genes play a pivotal role in vertebrates (Houweling et al., 2001; Lecaudey et al., 2005). the regulation of Notch signalling during Xenopus neural crest During neural development, Irx proteins act as repressors and formation (Glavic et al., 2004) and during Drosophila eye downregulate Bmp signalling to allow neural plate formation development (Domínguez and de Celis, 1998). The relationship (Gómez-Skarmeta et al., 2001; Itoh et al., 2002). In this work we between Irx genes and Notch signalling is currently under show that Irx proteins act as activators during kidney formation.

investigation. Thus, one possible mechanism of action of Irx proteins could be to DEVELOPMENT 3206 RESEARCH ARTICLE Development 135 (19) upregulate Bmp signalling, which is known to participate at many gastrula/early neurula stages, but not later. This developmental steps during vertebrate kidney formation (Cain et al., 2008). period is when RA is required for the activation of the early kidney Consistent with this idea, the reduction of Bmp signalling during genes Lim1 and Pax8 (Cartry et al., 2006), and it is well before Irx Xenopus neurula stages causes defects similar to those produced by genes are initially expressed. Thus, the RA effect on Irx gene Irx gene impairment (Bracken et al., 2008). Furthermore, we show expression is likely to be indirect, probably through Lim1 and Pax8. that increasing Bmp signalling partially rescue the kidney defects In addition, as we do not detect alteration of the Irx expression observed in Irx morphant and that Bmp7 expression is patterns when RA signalling is modulated just before proximal- downregulated in Irx-deficient embryos. Further experiments are distal patterning, RA influence on proximal-distal patterning required to determine more precisely the interaction between the (Wingert et al., 2007) is likely to occur as an indirect consequence Bmp pathway and Irx genes. of its effect on early kidney genes such as Lim1 and Pax8 (Cartry et al., 2006). Irx genes are required at later stages for In a functional survey of the enhancer activity of highly conserved proximal-distal pronephric patterning non-coding elements present in the IrxB complex (de la Calle- Recently, it has been reported that, within the pronephros field, Irx Mustienes et al., 2005) we identified two ultraconserved non-coding genes are initially expressed at tailbud stages and that only Irx3 is regions (UCRs) that activate expression in the pronephros, as required for proximal-distal pronephric patterning (Reggiani et al., determined in Xenopus transgenic experiments (de la Calle- 2007). As indicated above, we detect an earlier expression (at Mustienes et al., 2005). Each UCR is located in each Irx gene cluster neurula stage) of Irx1, Irx2 and Irx3 in that territory, and a in relatively close proximity to Irx1 and Irx3. These regulatory requirement for both Irx1 and Irx3 for the proper development of regions are likely to contribute to the regulation of Irx genes during the kidney territory before pronephric morphogenesis. As it was kidney development by early kidney specification genes. The possible that the late Irx3 described function might be an indirect detailed analyses of these regions should help unravel the molecular effect of the earlier Irx requirement, we have re-examined the mechanisms that control Irx gene expression during pronephros participation of Irx genes in proximal-distal patterning of the formation and to define the genetic cascade that operates during this pronephros. Although MOs are very useful reagents to reduce gene process. activity, their injections into blastomeres may impair gene function from early stages and make it difficult to recognize a late We are most grateful to E. Amaya, A. Brändli, E. Bellefroid, B. Blumberg, A. Fainsod, P. A. Krieg, D. Kimelman, N. Papalopulu, H. Sive, M. Taira, D. Turner, requirement. To try to uncouple early and late Irx requirements, we R. Vignali and P. Vize for reagents. We also thank N. Ueno and the NIBB/NIG injected low doses of Irx MOs. In these hypomorphic conditions, Xenopus laevis EST project for the Mochii clone XL056l08. We especially thank embryos did not show early kidney phenotypes but the late F. Casares, P. Lemaire, John Pearson and J. Modolell for helpful criticisms. This pronephric segmentation was affected. This suggests that both work was supported by grants from the Spanish Ministry of Education and genes are required for this late process but does not exclude that this Science (BFU2004-00310, BFU2007-60042/BMC, CSD2007-00008) and Junta de Andalucía (Proyecto de Excelencia 00260) to J.L.G.-S., and a Marie Curie could be an indirect consequence of an early requirement for Irx Reintegration Grant (ERG-014806) and a UPO Grant (APP2D06060) to P.A. genes. To determine Irx protein function unambiguously during E.R.-S. and P.A. are I3P fellows from the CSIC. proximal-distal pronephric patterning, we have used conditional loss- and gain-of-function of Irx genes. Overexpression of an Supplementary material Supplementary material for this article is available at inducible dominant-negative construct demonstrates that early http://dev.biologists.org/cgi/content/full/135/19/3197/DC1 impairment of Irx gene activity downregulates all proximal-distal markers examined. By contrast, late impairment of Irx activity References prevents only intermediate tubule formation. We confirmed this Blumberg, B., Bolado, J. J., Moreno, T. A., Kintner, C., Evans, R. M. and Papalopulu, N. (1997). An essential role for retinoid signaling in anteroposterior dual Irx protein function by overexpressing Irx1 or Irx3 at neurula neural patterning. Development 124, 373-379. or tailbud stages. Early Irx activation expands or promotes ectopic Bracken, C. M., Mizeracka, K. and McLaughlin, K. A. (2008). Patterning the expression of different segment markers, whereas late embryonic kidney: BMP signaling mediates the differentiation of the pronephric overexpression expands only the intermediate tubule marker Nkcc2. tubules and duct in Xenopus laevis. Dev. Dyn. 237, 132-144. Cain, J. E., Hartwig, S., Bertram, J. F. and Rosenblum, N. D. (2008). Bone Thus, our study reveals an earlier Irx gene requirement for most of morphogenetic protein signaling in the developing kidney: present and future. the phronephric field and also, in agreement with Reggiani et al. Differentiation doi:10.1111/j.1432-0436.2008.00265.x. (Reggiani et al., 2007), a late requirement, although in contrast to Carroll, T., Wallingford, J., Seufert, D. and Vize, P. D. (1999). Molecular regulation of pronephric development. Curr. Top. Dev. Biol. 44, 67-100. this report, our data support the necessity for both Irx1 and Irx3 for Carroll, T. J. and Vize, P. D. (1999). Synergism between Pax-8 and lim-1 in proximal-distal patterning. embryonic kidney development. Dev. Biol. 214, 46-59. Cartry, J., Nichane, M., Ribes, V., Colas, A., Riou, J. F., Pieler, T., Dolle, P., Bellefroid, E. J. and Umbhauer, M. (2006). Retinoic acid signalling is required Irx genes are regulated by RA signalling for specification of pronephric cell fate. Dev. Biol. 299, 35-51. RA is required for the activation of the early kidney genes Lim1 and Coffman, C. R., Skoglund, P., Harris, W. A. and Kintner, C. R. (1993). Pax8 (Cartry et al., 2006), and for the late regionalization of the Expression of an extracellular deletion of Xotch diverts cell fate in Xenopus pronephros (Wingert et al., 2007). We observe that the expansion of embryos. Cell 73, 659-671. de la Calle-Mustienes, E., Glavic, A., Modolell, J. and Gómez-Skarmeta, J. L. the kidney field associated with overexpression of Irx genes are (2002). Xiro homeoproteins coordinate cell cycle and primary neuron fomation similar to that found upon increasing retinoic acid activity (Cartry et by upregulating neuronal-fate repressors and downregulating the cell-cycle al., 2006). This suggests a possible link between Irx genes and RA inhibitor XGadd45-γ. Mech. Dev. 119, 69-80. de la Calle-Mustienes, E., Feijoo, C. G., Manzanares, M., Tena, J. J., signalling. Indeed, we find that both Irx1 and Irx3 are activated by Rodríguez-Seguel, E., Letizia, A., Allende, M. L. and Gómez-Skarmeta, J. RA signalling. This is consistent with the fact that the pronephric L. (2005). A functional survey of the enhancer activity of conserved non-coding expression of the gene encoding the RA producing enzyme Raldh2 sequences from vertebrate Iroquois cluster gene deserts. Genome Res. 15, and the RA receptor RAR precedes that of the Irx genes. Thus, RA 1061-1072. α Domínguez, M. and de Celis, J. F. (1998). A dorsal/ventral boundary established probably lies upstream of Irx genes during pronephros development. by Notch controls growth and polarity in the Drosophila eye. Nature 396, 276-

We also find that RA is necessary for Irx gene expression at late 278. DEVELOPMENT Irx genes in Xenopus kidney development RESEARCH ARTICLE 3207

Dressler, G. R. (2006). The cellular basis of kidney development. Annu. Rev. Cell Lecaudey, V., Anselme, I., Dildrop, R., Ruther, U. and Schneider-Maunoury, Dev. Biol. 22, 509-529. S. (2005). Expression of the zebrafish Iroquois genes during early nervous system Dudley, A. T., Lyons, K. M. and Robertson, E. J. (1995). A requirement for bone formation and patterning. J. Comp. Neurol. 492, 289-302. morphogenetic protein-7 during development of the mammalian kidney and Luo, G., Hofmann, C., Bronckers, A. L., Sohocki, M., Bradley, A. and eye. Genes Dev. 9, 2795-2807. Karsenty, G. (1995). BMP-7 is an inducer of nephrogenesis, and is also required Glavic, A., Gómez-Skarmeta, J. L. and Mayor, R. (2001). Xiro-1 controls for eye development and skeletal patterning. Genes Dev. 9, 2808-2820. mesoderm patterning by repressing Bmp-4 expression in the Spemann McLaughlin, K. A., Rones, M. S. and Mercola, M. (2000). Notch regulates cell Organizer. Dev. Dyn. 222, 368-376. fate in the developing pronephros. Dev. Biol. 227, 567-580. Glavic, A., Silva, F., Aybar, M. J., Bastidas, F. and Mayor, R. (2004). Interplay Reggiani, L., Raciti, D., Airik, R., Kispert, A. and Brandli, A. W. (2007). The between Notch signaling and the homeoprotein Xiro1 is required for neural prepattern Irx3 directs nephron segment identity. Genes Dev. crest induction in Xenopus embryos. Development 131, 347-359. 21, 2358-2370. Gómez-Skarmeta, J. L. and Modolell, J. (2002). iroquois genes: genomic Ryffel, G. U. (2003). What can a frog tell us about human kidney development. organization and function in vertebrate neural development. Curr. Opin. Genet. Nephron Exp. Nephrol. 94, e35-e43. Dev. 12, 403-408. Seufert, D. W., Brennan, H. C., DeGuire, J., Jones, E. A. and Vize, P. D. (1999). Gómez-Skarmeta, J. L., de la Calle-Mustienes, E., Modolell, J. and Mayor, R. Developmental basis of pronephric defects in Xenopus body plan phenotypes. (1999). Xenopus brain factor-2 controls mesoderm, forebrain and neural crest Dev. Biol. 215, 233-242. development. Mech. Dev. 80, 15-27. Taelman, V., Van Campenhout, C., Solter, M., Pieler, T. and Bellefroid, E. J. Gómez-Skarmeta, J. L., de la Calle-Mustienes, E. and Modolell, J. (2001). The (2006). The Notch-effector HRT1 gene plays a role in glomerular development and Wnt-activated Xiro1 gene encodes a repressor that is essential for neural patterning of the Xenopus pronephros anlagen. Development 133, 2961-2971. development and dowregulates BMP4. Development 128, 551-560. Tena, J. J., Neto, A., de la Calle-Mustienes, E., Bras-Pereira, C., Casares, F. Harland, R. (1991). In situ hybridization: an improved whole mount method for and Gomez-Skarmeta, J. L. (2007). Odd-skipped genes encode repressors that Xenopus embryos. Methods Cell Biol. 36, 685-695. control kidney development. Dev. Biol. 301, 518-531. Harland, R. and Weintraub, H. (1985). Translation of mRNA injected into Turner, D. L. and Weintraub, H. (1994). Expression of achaete-scute homolog 3 Xenopus oocytes is specifically inhibited by antisense RNA. J. Cell Biol. 101, in Xenopus embryos converts ectodermal cells to a neural fate. Genes Dev. 12, 1094-1099. 1434-1447. Heller, N. and Brandli, A. W. (1999). Xenopus Pax-2/5/8 orthologues: novel Vize, P. D., Jones, E. A. and Pfister, R. (1995). Development of the Xenopus insights into Pax gene evolution and identification of Pax-8 as the earliest marker pronephric system. Dev. Biol. 171, 531-540. for otic and pronephric cell lineages. Dev. Genet. 24, 208-219. Wang, S., Krinks, M., Kleinwaks, L. and Moos, M., Jr (1997). A novel Xenopus Houweling, A. C., Dildrop, R., Peters, T., Mummenhoff, J., Moorman, A. F. homologue of bone morphogenetic protein-7 (BMP-7). Genes Funct. 1, 259-271. M., Rüther, U. and Christoffels, V. M. (2001). Gene and cluster-specific Wingert, R. A., Selleck, R., Yu, J., Song, H. D., Chen, Z., Song, A., Zhou, Y., expression of the Iroquois family members during mouse development. Mech. Thisse, B., Thisse, C., McMahon, A. P. et al. (2007). The cdx genes and Dev. 107, 169-174. retinoic acid control the positioning and segmentation of the zebrafish Itoh, M., Kudoh, T., Dedekian, M., Kim, C. H. and Chitnis, A. B. (2002). A role pronephros. PLoS Genet. 3, 1922-1938. for iro1 and iro7 in the establishment of an anteroposterior compartment of the Zhou, X. and Vize, P. D. (2004). Proximo-distal specialization of epithelial ectoderm adjacent to the midbrain-hindbrain boundary. Development 129, transport processes within the Xenopus pronephric kidney tubules. Dev. Biol. 2317-2327. 271, 322-338. DEVELOPMENT